In spring 2024, 70 researchers from 14 countries worked together over the course of 72 hours to build an MRI instrument from scratch. The “design-build-test” hackathon (named “ezyMRI”), hosted at Singapore University of Technology and Design, began with one day of seminars and planning, followed by two days and a half of construction, and wrapped up with a final half-day dedicated to system integration and testing. Participants – many of them graduate students and early-career researchers, and some with little or no prior MRI experience – were divided into six multidisciplinary teams focusing on magnet, gradient coil, RF coil, console, system integration, and industrial design. Together, they produced a working instrument capable of capturing phantom images before the clock ran out.
“There’s growing momentum behind open-source and DIY approaches to scientific instrumentation,” says Shaoying Huang, who organized the event. “This is fueled by advances in affordable additive manufacturing (e.g., 3D printing for fast prototyping) and high-performance personal computers.” Advances in low-cost hardware were front and center at the event: the magnet team used ferrite permanent magnets in a Halbach configuration for portability; the gradient coil team designed a setup with an ~8 cm field of view, supported by 3D-printed parts for fast assembly and precise positioning; the RF coil team constructed a transmit/receive solenoid coil, tuned and matched on-site; and the console team adapted open-source, FPGA-based electronics from earlier low-field MRI projects.
The biggest challenge, Huang recalls, came in the lead-up to the event. “There was a long list of tools, equipment, and components that we needed to prepare. The construction of the sub-systems and the system assembly also needed to be rehearsed.” During the build, there were moments of creative problem solving – and brute force. “The magnet team finished assembling a new array by working long hours each day; the gradient coil team used a soldering iron to assemble 3D-printed pieces to form a complete housing. There was a slight mismatch of the dimensions of the magnet and the gradient coils – luckily we successfully hammered the gradient coils into the magnet.” The paper also notes the use of improvised spacer rings to fine-tune alignment.
The cost of the basic ingredients and components for the prototype came in at only a few thousand US dollars – illustrating the potential of open-source design for affordable medical imaging. Huang believes 3D printing could further progress portable MRI hardware research in the lab in the near future. “It’s good for fast prototyping and allows more design possibilities for magnet positioning and for wire positioning for both RF coils and gradient coils. With that, it may allow hardware simplifications, which is critical for portable MRI.”
Although this rapid, collaborative prototyping approach could be applied to other analytical instruments, such as spectrometers, Huang suspects it might not have the same level of societal impact. For low-field MRI, however, the implications are profound. Huang is optimistic about its potential in global health and field settings – but stresses that the hackathon’s main goal was educational. Post-event surveys showed a marked increase in participants’ confidence in MRI hardware design, and several have since launched follow-up projects using the open-source designs. “If we want to grow the field fast so portable MRI scanners can reach a wide population, we need a lot of capable engineers and researchers to research, develop, and maintain the systems,” she says. “With that thought, I believe a hackathon like that can grow interest in MRI hardware building among young researchers and engineers.”
Newsletters
Receive the latest analytical science news, personalities, education, and career development – weekly to your inbox.
